¹ A number of reasons may be given for the apparent shift in opinion: (1) George Bush's two conservative appointees, John Roberts and Samuel Alito (‘Justices Back Ban on Method of Abortion’, New York Times, 19 April 2007 (<nytimes.com/2007/04/19/Washington/19scotus.html>)); (2) the apparent replacement of Sandra Day O’Connor with either Anthony Kennedy or Samuel Alito as the swing vote (‘The Supreme Court: Conservatives resurgent’, The Economist, 21 April 2007, p. 54; Joannan Grossman and Linda McClain, ‘New Justices, New Rules: The Supreme Court Upholds the Federal Partial-Birth Abortion Ban Act of 2003’, Findlaw: Legal News and Commentary (1 May 2007), <https://writ.news.findlaw.com/commentary/20070501mcclain.html>); (3) the Court following the opinion of an American public which favours some restrictions on abortion (Fiorina, Morris P., Abrams, Samuel J. and Pope, Jeremy C., Culture War? The Myth of a Polarized America (New York: Pearson Longman, 2005), pp. 42–4Google Scholar; John Yoo, ‘Partial-birth bigotry’, Wall Street Journal, 28 April 2007, republished by the American Enterprise Institute, 30 April 2007 (<aei.org>)); and (4) the emergence of a Catholic majority among Supreme Court justices (‘The Supreme Court's Catholic majority’, New York Times, 25 April 2007 (<nytimes.com/2007/04/25/us/politics>); ‘Subtext of abortion ruling: Religion’, International Herald Tribune, 26 April 2007, p. 4; Yoo, ‘Partial-birth bigotry’).

² ‘Partial-birth Abortion Ban Act of 2003’, 18 USC, 1531, Public Law 108–105.

³ In a similar vein, Scott L. Feld, Bernard Grofman and Joseph Godfrey, ‘Putting a Spin on It: Geometric Insights into How Candidates with Seemingly Losing Positions Can Still Win’ (unpublished manuscript, 2006) argue that politicians may use ‘spin’ (or the reframing of issues by use of persuasion) to simplify multiple issue dimensions into a single new dimension, and thereby improve their competitive edge among voters.

⁴ Poole, Keith T., Spatial Models of Parliamentary Voting (Cambridge: Cambridge University Press, 2005).CrossRefGoogle Scholar

⁵ A number of recent studies of legislative and other political behaviour have sought to explore the links between multidimensionality and low-dimensionality. Some of these explore the question by analysing roll-call votes (Groseclose, Tim, Levitt, Steven D. and Snyder, James M. Jr, ‘Comparing Interest Group Scores Across Time and Chambers: Adjusted ADA Scores for the U.S. Congress’, American Political Science Review, 93 (1990), 33–50CrossRefGoogle Scholar; Norton, Noelle H., ‘Uncovering the Dimensionality of Gender Voting in Congress’, Legislative Studies Quarterly, 24 (1999), 65–86CrossRefGoogle Scholar; Hurwitz, Mark S., Moiles, Roger J. and Rohde, David W., ‘Distributive and Partisan Issues in Agriculture Policy in the 104th House’, American Political Science Review, 95 (2001), 911–22)CrossRefGoogle Scholar; others by analysing survey or elite interview data with correlational and confirmatory factor analysis (Craig, Stephen C., Martinez, Michael D. and Kane, James G., ‘The Structure of Political Competition: Dimensions of Candidate and Group Evaluation Revisited’, Political Behavior, 21 (1999), 283–304Google Scholar; Levine, Jeffrey, Carmines, Edward G. and Sniderman, Paul M., ‘The Empirical Dimensionality of Racial Stereotypes’, Public Opinion Quarterly, 63 (1999), 371–84Google Scholar) or principal-component analysis (Selck, Torsten J., ‘On the Dimensionality of European Union Legislative Decision-making’, Journal of Theoretical Politics, 16 (2004), 203–22CrossRefGoogle Scholar); and others by comparing and contrasting alternative methodological approaches (Brazill, Timothy J. and Grofman, Bernard, ‘Factor Analysis versus Multidimensional Scaling: Binary Choice Roll-Call Voting and the U.S. Supreme Court’, Social Networks, 24 (2002), 201–29CrossRefGoogle Scholar; Grofman, Bernard and Brazill, Timothy J., ‘Identifying the Median Justice on the Supreme Court Through Multidimensional Scaling: Analysis of “Natural Courts” 1953–1991’, Public Choice, 112 (2002), 55–79)CrossRefGoogle Scholar. Two further studies offer insights on multi- and low-dimensionality by examining the legislative process, both before and after floor voting (Potoski, Matthew and Talbert, Jeffery, ‘The Dimensional Structure of Policy Outputs: Distributive Policy and Roll Call Voting’, Political Research Quarterly, 53 (2000), 695–710CrossRefGoogle Scholar; Ringe, Nils, ‘Policy Preference Formation in Legislative Politics: Structures, Actors, and Focal Points’, American Journal of Political Science, 49 (2005), 731–45)Google Scholar.

⁶ My thanks to Albert Weale for helping to clarify this point.

⁷ Kingdon, John W., Congressmen's Voting Decisions, 3rd edn (Ann Arbor: University of Michigan Press, 1989)Google Scholar; Mayhew, David R., Congress: The Electoral Connection (New Haven, Conn.: Yale University Press, 1974)Google Scholar; Fenno, Richard, Congressmen in Committtees (Boston, Mass.: Little, Brown, 1973)Google Scholar; Uslaner, Eric M., The Movers and the Shirkers: Representatives and Ideologues in the Senate (Ann Arbor: University of Michigan Press, 1999)CrossRefGoogle Scholar; Poole, Keith T. and Rosenthal, Howard, Congress: A Political-Economic History of Roll-Call Voting (Oxford: Oxford University Press, 1997)Google Scholar.

⁸ Less than one hour after the ban became law, Richard Kopf, Chief Judge of the US District Court for Nebraska, issued a limited temporary restraining order against the ban, arguing that it lacked an exception if the woman's health is at risk. The Act was subsequently ruled unconstitutional in other federal district courts. In 2004, the four doctors who had challenged the Nebraska state law in Stenberg v. Carhart disputed the congressional findings, arguing that the procedure is safe under some conditions and could avoid complications to the woman. In Carhart v. Ashcroft (2004), Judge Kopf ruled that the four doctors had demonstrated that the congressional findings were ‘unreasonable’ and that the Act was unconstitutional on the grounds that it failed to provide a health exception for women. In 2005, the 8th Circuit Court in St Louis upheld the decision of Kopf in Carhart v. Ashcroft, whereupon the Bush Administration appealed (Gonzales v. Carhart, No. 05–380). In 2006, the Supreme Court agreed to hear the appeal.

⁹ Martin, Andrew D., ‘Congressional Decision Making and the Separation of Powers’, American Political Science Review, 95 (2001), 361–78.CrossRefGoogle Scholar

¹⁰ Morgan, Donald G., Congress and the Constitution: A Study of Responsibility (Cambridge, Mass.: Belknap Press of Harvard University Press, 1966)CrossRefGoogle Scholar; Whittington, Keith E., Constitutional Construction: Divided Powers and Constitutional Meaning (Cambridge, Mass.: Harvard University Press, 1999)Google Scholar.

¹¹ Quirk, Paul J., ‘Deliberation and Decision Making’, in Quirk, Paul J. and Binder, Sarah A., eds, The Legislative Branch (Oxford: Oxford University Press, 2005), pp. 314–48Google Scholar; and Mucciaroni, Gary and Quirk, Paul J., Deliberative Choices: Debating Public Policy in Congress (Chicago: University of Chicago Press, 2006)Google Scholar.

¹² One need look no further than contemporary media coverage to sustain this assessment. However, a number of recent books on abortion politics and abortion discourse in the United States have sought to explain the underlying conflicts between the pro-life and pro-choice camps, including: Marx Ferree, Myra, Anthony Gamson, William, Gerhards, Jürgen and Rucht, Dieter, Shaping Abortion Discourse: Democracy and the Public Sphere in Germany and the United States (Cambridge: Cambridge University Press, 2002)Google Scholar; Maxwell, Carol J., Pro-Life Activists in America: Meaning, Motivation and Direct Action (Cambridge: Cambridge University Press, 2002)Google Scholar; Saletan, William, Bearing Right: How Conservatives Won the Abortion War (Berkeley: University of California Press, 2004)Google Scholar; Burns, Gene, The Moral Veto: Framing Contraception, Abortion, and Cultural Pluralism in the United States (Cambridge: Cambridge University Press, 2005)CrossRefGoogle Scholar.

¹³ Adams, Greg D., ‘Abortion: Evidence of an Issue Evolution’, American Journal of Political Science, 41 (1997), 718–37CrossRefGoogle Scholar.

¹⁴ Poole and Rosenthal, pp. 112–13.

¹⁵ The terms ‘pro-choice’ and ‘pro-life’ are clearly value-laden and may thus be considered objectionable. Yet inasmuch as these terms are commonly used in both popular and academic works, and they embody the frames used by both camps of the abortion debate, they are used here – although their problematic nature is acknowledged.

¹⁶ Carmines and Wood refine the issue evolution thesis by confirming that activists in both parties have followed this pattern, with Democratic activists moving towards pro-choice and Republican activists towards pro-life (Carmines, Edward G. and Woods, James, ‘The Role of Party Activists in the Evolution of the Abortion Issue’, Political Behavior, 24 (2002), 361–77Google Scholar). These authors find, moreover, that from the 1990s onwards, abortion attitudes in the electorate have followed the polarization pattern set by political elites. These authors and Adams identify a trend towards party polarization on abortion beginning in Congress and subsequently spreading to the electorate. Fiorina et al., by contrast, take issue with the broader notion of a polarized political culture in the United States.Onabortion, Americans are said to be ‘pro-choice, buts’ – that is, in favour of abortion with some restrictions (Morris P. Fiorina, et al., Culture War? pp. 42–4). With respect to party polarization on abortion, these authors readily accept a clear divide between pro-choice Democrats and pro-life Republicans who are party activists, political office-holders and lobbyists groups. Contrary to Carmines and Woods, they argue that rank-and-file Democrats and Republicans have not absorbed the signals from activists and politicians on abortion. So, while all these authors accept a trend in party polarization on abortion among political elites and activists, researchers continue to debate the extent to which the mass public has followed this clear trend (for example, Carsey, Thomas M. and Layman, Geoffrey C., ‘Changing Sides or Changing Minds? Party Identification and Policy Preferences in the American Electorate’, American Journal of Political Science, 50 (2006), 464–77CrossRefGoogle Scholar) and whether rank-and-file Democrats and Republicans have listened to the partisan message from the elites and activists.

¹⁷ Martin Haskell, ‘Dilation and Extraction for Late Second Trimester Abortion’ (National Abortion Federation Risk Management Seminar, Dallas, Texas, 1992).

¹⁸ Haskell, ‘Dilation and Extraction for Late Second Trimester Abortion’, pp. 30–1.

¹⁹ Saletan, Bearing Right, p. 233.

²⁰ Saletan, Bearing Right, p. 233.

²¹ PP – Planned Parenthood Federation, ‘Abortion after the First Trimester: Legality of Abortion (Laws and Specific Abortion Techniques)’, <http://www.plannedparenthood.org/pp2/portal/> accessed 18 March 2006.

²² Jason Abaluck, ‘Partial truths: The partial-birth abortion ban violates women's rights’, Perspective, Harvard-Radcliffe's Liberal Monthly (November 2003), <http://www.digitas.harvard.edu/~perspy/issues/2003/nov/partial.html>, accessed 24 March 2006.

²³ The president may reject (or veto) a bill passed by both houses of Congress. In rare circumstances, Congress may override a presidential veto, in which case the bill becomes law without the approval of the president. A two-thirds majority vote of the representatives and senators present is required to override a presidential veto.

²⁴ Partial-birth Abortion Ban Act of 2003, 18 USC, 1531, 108–105, sec. 2 (1, 2).

²⁵ Sen. Boxer and other opponents no doubt sought, at the minimum, to revise the bill in committee or stall its progress.

²⁶ A conference committee reconciles differences between the House and Senate versions of a bill. Once a compromise version is agreed by the committee, it prepares a report which details the proposed changes. Both the House and Senate must approve this report.

²⁷ Poole and Rosenthal, Congress, p. 113.

²⁸ Poole, Spatial Models of Parliamentary Voting, pp. 147–8.

²⁹ Poole, Spatial Models of Parliamentary Voting, p. 48.

³⁰ Keith Poole, ‘108th Senate Rank Ordering’, (26 October 2004) <http://www.voteview.com/>, accessed January 2006.

³¹ Poole, ‘108th Senate Rank Ordering’.

³² The debates were partially edited to exclude the frequent requests for time, statements by the clerk and the presiding officer, the reading of the bill and the calling of the roll.

³³ Gabel, Matthew J. and Huber, John D., ‘Putting Parties in Their Place: Inferring Party Left–Right Ideological Positions from Party Manifestos Data’, American Journal of Political Science, 44 (2000), 94–103CrossRefGoogle Scholar; Laver, Michael and Garry, John, ‘Estimating Policy Positions from Political Texts’, American Journal of Political Science, 44 (2000), 619–34CrossRefGoogle Scholar; Laver, Michael and Benoit, Kenneth, ‘Locating TDs in Policy Spaces Using Computer Word-scoring’, Irish Political Studies, 17 (2002), 59–72CrossRefGoogle Scholar; Laver, Michael, Benoit, Kenneth and Garry, John, ‘Estimating the Policy Positions of Political Actors Using Words as Data’, American Political Science Review, 97 (2002), 311–31Google Scholar; Kenneth Benoit and Slava Mikhailov, ‘Correcting the Error in the Comparative Manifesto Project Estimates’ (paper presented at the Annual Meeting of the Midwest Political Science Association, Chicago, 2007); Jeremy Albright, ‘Political Parties in Time and Space’ (paper presented at the Annual Meeting of the Midwest Political Science Association, Chicago, 2007); and Jonathan Slapin and Sven-Oliver Proksch, ‘A Scaling Model for Estimating Time-Series Policy Positions from Texts’ (paper presented at the Annual Meeting of the Midwest Political Science Association, Chicago, 2007).

³⁴ Kevin M. Quinn et al., ‘An Automated Method of Topic-Coding Legislative Speech Over Time with Application to the 105th–108th U.S. Senate’ (unpublished manuscript, 18 July 2006).

³⁵ G. David Garson, ‘Researching and Teaching Political Culture through Web-Based Content Profile Analysis’ (paper presented at the Annual Meeting of the American Political Science Association, Boston, Mass., 2002).

³⁶ See the following papers presented at the Annual Meeting of the Midwest Political Science Association, Chicago, 2007: Daniel Hopkins and Gary King, ‘Extracting Systematic Social Science Meaning from Text’; Dustin Hillard, Stephen Purpura and John Wilkerson, ‘An Active Learning Framework for Classifying Political Text’; Burt Monroe, Kevin Quinn and Michael Colaresi, ‘Legislative Rhetoric and Heresthetics’; Jean-François Godbout, Daniel Diermeier, Bei Yu and Stefan Kaufman, ‘Automated Text Classification of Senatorial Speech in the 101st–109th Congresses for Ideological Vocabulary Extraction and Position Prediction’.

³⁷ For a recent showcase of these approaches in political science, see <http://www.purpuras.net/apsagroup/>.

³⁸ Will Lowe, ‘Understanding Wordscores” (paper presented at the Annual Meeting of the Midwest Political Science Association, Chicago, 2007).

³⁹ Some examples include Laver et al., ‘Estimating the Policy Positions of Political Actors’, and Hopkins and King, ‘Extracting Systematic Social Science Meaning from Text’.

⁴⁰ Dustin Hillard et al., ‘An Active Learning Framework for Classifying Political Text’.

⁴¹ Noel-Jorand, M. C., Reinert, M., Bonnon, M. and Therme, P., ‘Discourse Analysis and Psychological Adaptation to High Altitude Hypoxia’, Stress Medicine, 11 (1995), 27–39CrossRefGoogle Scholar; Noel-Jorand, M. C., Reinert, M., Giudicelli, S. and Dassa, D., ‘A New Approach to Discourse Analysis in Psychiatry, Applied to Schizophrenic Patient Speech’, Schizophrenia Research, 25 (1997), 183–98Google Scholar; Noel-Jorand, M. C., Reinert, M., Giudicelli, S. and Dassa, D., ‘Schizophrenia: The Quest for a Minimum Sense of Identity to Ward Off Delusional Psychosis’, Canadian Journal of Psychiatry, 49 (2004), 394–8CrossRefGoogle Scholar; Lahlou, L., ‘A Method to Extract Social Representations from Linguistic Corpora’, Japanese Journal of Experimental Social Psychology, 36 (1996), 278–91Google Scholar; Nicholas C. Allum, ‘A Social Representations Approach to the Comparison of Three Textual Corpora Using ALCESTE’ (MSc dissertation, London School of Economics and Political Science, 1998); Mathieu Brugidou, ‘Epitaphes, l’image de François Mitterrand à travers l’analyse d’une question ouverte posée à sa mort’ (Epitaphs. Francois Mitterrand's Image: An Analysis of an Open Question Asked on His Death), Revue Française de Science Politique, 48 (1998), 97–120; Mathieu Brugidou, ‘Les discours de la revendication et de l’action dans les éditoriaux de la presse syndicale (1996–1998)’ (The Discourse of Demands and Action in [French] Trade Union Press Editorials (1996–1998)), Revue Française de Science Politique, 50 (2000), 967–92; Andrew Bailey and Cheryl Schonhardt-Bailey, ‘Explaining the Volcker Revolution of 1979: Testing Theories with Transcripts’ (paper presented at the Annual Meeting of the American Political Science Association, Washington, D.C., 2005); Andrew Bailey and Cheryl Schonhardt-Bailey, ‘Policy Shaping Politics; Monetary Policy Deliberations in Congressional Hearings’ (paper presented at the First World Meeting of the Public Choice Society, 2007); Schonhardt-Bailey, Cheryl, From the Corn Laws to Free Trade: Interests, Ideas and Institutions in Historical Perspective (Cambridge: MIT Press, 2006).Google Scholar

⁴² Although subsequent versions may allow a larger corpus, ALCESTE 4.7 requires that the corpus not exceed 15 mb.

⁴³ Plurals and conjugation endings are reduced to a single form and nonce words are eliminated from the analysis. This leaves a smaller word count which is analysed by the program.

⁴⁴ These are deemed ‘passive’ as they do not contribute to either the calculation of the word classes or the factors in the correspondence analysis.

⁴⁵ Popping notes that the ECU is akin to the ‘recording unit’ used in other programs, where it is usually defined by the researcher (Roel Popping, correspondence with author, 29 October 2004).

⁴⁶ A contextual unit is equivalent to one or more successive ECU(s). The two calculations are done with two different parameters for the selected number of words per contextual unit in order to check the reliability of the classes and the stability of the results (Max Reinert, ALCESTE Users’ Manual, 4.0 Pro (English version) (Toulouse: Image, 1998).

⁴⁷ The standard report lists the top twenty ECUs for each class, ranked by x ² association. However, a separate file is produced that lists all the ECUs for each class, where the default cut-off for selection is 0.

⁴⁸ Meaningful words are nouns, verbs, adjectives and some adverbs. These comprise ‘the vocabulary of the corpus’ and are used to calculate the classes. Function words (because, where, as and so on) are recognized by the internal dictionary and are required for syntax, but – like the tagged indicators – are not used in the calculation of classes. (Reinert, ALCESTE Users’ Manual, p. 12) For a table of the levels of x ² values for words, which may then be used in conjunction with a standard x ² table, see Reinert, ALCESTE Users’ Manual, p. 41.

⁴⁹ Using descending hierarchical classification, ALCESTE creates two binary tables (or two classifications) which are then compared. The ECUs that are retained for analysis are those found in both classifications – a process which effectively creates a stable level of partitioning. The stability of the partitioning is measured by constructing a table of co-occurrence, which is the cross between the partition obtained in the first classification and the partition obtained in the second classification. The result is a ‘signed x ² table’ – that is a data table with the positive and negative links between the classes. This signed table is used to test the statistical significance of the table of co-occurrence, where ‘a high positive chi-square indicate(s) a strong positive link (a chance occurrence would be expected to be less), a high negative chi-square, a strong negative link (a chance occurrence would be expected to be more)’ ( Reinert, ALCESTE Users’ Manual, pp. 26–7).

⁵⁰ This minimum value for word selection within ALCESTE varies from 2.13 to 20, with smaller text files tending towards the lower threshold and larger ones towards the high threshold. The basic rule of thumb with ALCESTE is (as with any statistical analysis) – the more data, the easier it is to attain statistical significance (hence larger text files have to attain a higher threshold to be statistically significant).

⁵¹ Burns, The Moral Veto, pp. 10–11.

⁵² Morgan, Congress and the Constitution, p. 11.

⁵³ Burns, The Moral Veto, pp. 10–11.

⁵⁴ The growing literature on ‘framing’ examines the role of ideas in structuring choices (Druckman, James N., ‘On the Limits of Framing Effects: Who Can Frame?’ Journal of Politics, 63 (2001), 1041–66Google Scholar; Druckman, James N., ‘Political Preference Formation: Competition, Deliberation, and the (Ir)relevance of Framing Effects’, American Political Science Review, 98 (2004), 671–86CrossRefGoogle Scholar; Iyengar, Shanto, ‘Speaking of Values: The Framing of American Politics’, The Forum (Berkeley Electronic Press [bepress]), 3, 3, art. 7 (2005)Google Scholar; Johnston, Hank and Noakes, John A., Frames of Protest: Social Movements and the Framing Perspective (Lanham, Md.: Rowman & Littlefield, 2005))Google Scholar. Equating framing strategies with those of heresthetics, Druckman explains that both entail shifting the emphasis of a policy proposal from one set of beliefs to another set of beliefs (Druckman, ‘On the Limits of Framing Effects’, pp. 1044, 1046). Persuasion along a single dimension (or within a single frame) is quite different. An individual may be persuaded towards one or the other end of the dimension/frame, but this reflects a movement within a given spectrum of beliefs (for example, from liberalism to conservatism). Heresthetics, or framing effects, would entail a change of beliefs rather than a movement along a given spectrum.

⁵⁵ Greenacre, Michael and Hastie, Trevor, ‘The Geometric Interpretation of Correspondence Analysis’, Journal of the American Statistical Association, 82 (1987), 437–47CrossRefGoogle Scholar; Greenacre, Michael J., Correspondence Analysis in Practice (London: Academic Press, 1993)Google Scholar. While correspondence analysis is well established in the French literature (see Benzecri, J-P., L’Analyse des Données. Tome1: La Taxinomie. Tome 2: L’Analyse des Correspondances (Paris: Dunod, 1973)Google Scholar, and the journal Cahiers de l’Analyse des Données), its use has spread with the publication of English applications (Greenacre, Michael J. and Underhill, L. G., ‘Scaling a Data Matrix in Low-dimensional Euclidean Space’, in Hawkins, D.M., ed., Topics in Applied Multivariate Analysis (Cambridge: Cambridge University Press, 1982Google Scholar); Greenacre, Michael J., Theory and Applications of Correspondence Analysis (London: Academic Press, 1984)Google Scholar; and Weller, Susan C. and Romney, A. Kimball, Metric Scaling: Correspondence Analysis (London: Sage, 1990)CrossRefGoogle Scholar) and is occasionally used by political scientists (Blasius, Jorg and Thiessen, Victor, ‘Methodological Artifacts in Measures of Political Efficacy and Trust: A Multiple Correspondence Analysis’, Political Analysis, 9 (2001), 1–20)CrossRefGoogle Scholar. Correspondence analysis using numerical data is available in several major statistical packages, including BMDP, SPSS and SAS.

⁵⁶ Reinert, ALCESTE Users’ Manual, p. 45.

⁵⁷ For this, correspondence analysis uses the ‘x ² distance’, which resembles the Euclidean distance between points in physical space. (Here, x ² distance – which is distinct from the x ² statistic used to measure the significance of the words and tags – can be observed in Euclidean space by transforming the profiles before constructing the plots.) In correspondence analysis, each squared difference between co-ordinates is divided by the corresponding element of the average profile (where the profile is a set of frequencies divided by their total). The justification for using the x ² concept is that it allows one to transform the frequencies by dividing the square roots of the expected frequencies, thereby equalizing the variances. This can be compared to factor analysis, where data on different scales are standardized. For more detailed discussion and further geometric reasons for using the x ² distance in correspondence analysis, see Greenacre, Correspondence Analysis in Practice, pp. 34–6.

⁵⁸ Correspondence analysis usually refers to the ‘inertia’ of a table, which can also be called ‘association’ (Weller and Romney, Metric Scaling). A corresponding x ² value can be obtained by multiplying the association value by the total n of the table.

⁵⁹ The association and x ² statistic may be interpreted geometrically as the degree of dispersion of the set of rows and columns (or profile points) around their average, where the points are weighted.

⁶⁰ In total, three factors are identified in the correspondence analysis (with the third factor obtaining an eigenvalue of 0.16 and association of 22.7 per cent). (Usually, the dimensionality of the system is one less than the number of classes in the profile (Greenacre, Correspondence Analysis in Practice, p. 14).)

⁶¹ Quoted in Saletan, Bearing Right, p. 234.

⁶² As noted earlier, a further factor accounts for another 22.7 per cent of the association, but this is not explored in this article.

⁶³ Justice Kennedy, Gonzales, Attorney General v. Carhart et al., certiorari to the United States Court of Appeals for the Eight Circuit, Argued 8 November 2006 – Decided 18 April 2007, No. 05–380.

⁶⁴ Reinert, Max, ‘Une methode de classification descendante hierarchique: application a l’analyse lexicale par contexte’, Les Cahiers de l’Analyse des Données, 8 (1983), 187–98Google Scholar; Reinert, ALCESTE Users’ Manual; Reinert, Max, ‘Le rôle de la répétition dans la représentation du sens et son approche statistique dans la méthode ALCESTE’, Semiotica, 147 (2003), pp. 389–420Google Scholar.

⁶⁵ Bauer, Martin, ‘Classical Content Analysis: A Review’, in Bauer, Martin W. and Gaskell, George, eds, Qualitative Researching with Text, Image and Sound: A Practical Handbook (London: Sage, 2000), pp. 131–51CrossRefGoogle Scholar; Noel-Jorand et al., ‘Discourse Analysis and Psychological Adaptation to High Altitude Hypoxia’; Noel-Jorand et al., ‘A New Approach to Discourse Analysis in Psychiatry’; Noel-Jorand et al. ‘Schizophrenia’; Lahlou, L., ‘A Method to Extract Social Representations from Linguistic Corpora’, Japanese Journal of Experimental Social Psychology, 36 (1996), 278–91Google Scholar; Guerin-Pace, France, ‘Textual Statistics. An Exploratory Tool for the Social Sciences’, Population: An English Selection: New Methodological Approaches in the Social Sciences, 10 (1998), 73–95Google ScholarPubMed.

⁶⁶ Brugidou, Mathieu, ‘Epitaphes, l’image de François Mitterrand à travers l’analyse d’une question ouverte posée à sa mort’, Revue Française de Science Politique, 48 (1998), 97–120Google Scholar; Brugidou, Mathieu, ‘Argumentation and Values: An Analysis of Ordinary Political Competence Via An Open-Ended Question’, International Journal of Public Opinion Research, 15 (2003), 413–30Google Scholar; Schonhardt-Bailey, Cheryl, ‘Measuring Ideas More Effectively: An Analysis of Bush and Kerry's National Security Speeches’, PS: Political Science and Politics, 38 (2005), 701–11Google Scholar; Schonhardt-Bailey, Cheryl, From the Corn Laws to Free Trade: Interests, Ideas and Institutions in Historical Perspective (Cambridge, Mass.: MIT Press, 2006)CrossRefGoogle Scholar.

⁶⁷ Nicole Kronberger and Wolfgang Wagner, ‘Keywords in Context: Statistical Analysis of Text Features’, in Bauer and Gaskell, eds, Qualitative Researching with Text, Image and Sound, p. 306.

⁶⁸ Bertier, P. and Bouroche, J.M., Analyse des données multidimensionnelles (Paris: Presses Universitaires de France, 1975)Google Scholar; Hayashi, C., ‘On the Quantification of Qualitative Data from the Mathematics-Statistical Point of View’, Annals of the Institute of Statistical Mathematics, 2 (1950)CrossRefGoogle Scholar; Benzecri, J. P., Pratique de l’analyse des données: linguistique et lexicologie (Paris: Dunod, 1981)Google Scholar; Greenacre, Correspondence Analysis in Practice. Quotation from Kronberger and Wagner, ‘Keywords in Context: Statistical Analysis of Text Features’, p. 306.

⁶⁹ Roel Popping, correspondence with author, 29 October 2004.

⁷⁰ For ALCESTE, ‘statements’ are defined as ‘contextual units’. The program automatically determines contextual units with reference to punctuation and the length of the statement up to a maximum of 250 characters.

⁷¹ Kronberger and Wagner, ‘Keywords in Context’, p. 308.

⁷² Max Reinert, correspondence with author, 24 October 2006.

⁷³ Kronberger and Wagner, ‘Keywords in Context’, p. 309.

⁷⁴ Kronberger and Wagner, ‘Keywords in Context’.